Thoughts about the Democrat Donnybrook: Why Hillary Will Win on the 6th

I've had a number of lengthy email exchanges with blog readers and friends (most of them outside of the state of Indiana) who have expressed curiosity about the imminent Indiana primary and how this, the most Republican of states, will turn when it comes to a Democrat presidential primary. This post is an attempt to synthesize and combine those conversations into a coherent discussion of the Democrat presidential primary.

Warning; monster post follows.

Polling has been all over the place, as the Real Clear Politics average indicates. This fluidity has a lot more to do, I think, with differing polling methodologies than it does with actual fluidity in the race itself.

The most accurate poll was probably the recent SUSA poll that indicated a nine point lead by Clinton. I tend to distrust SUSA as a polling firm, but they're loads better than the Indy Star's polling firm. The Star's pollster did a general election poll back late last year whose sample was biased and unbalanced, containing more Democrats than Republicans. If you don't know that general election polls in Indiana should have more declared Republicans than Democrats, then you shouldn't be trusted to conduct polls here.

Indiana is also a difficult state to poll; there have been no seriously contested statewide primaries in Indiana in almost three decades. This absence of historical data makes determining turnout models and other weighting factors difficult (and primaries are hard enough to poll as it is).

This situation is complicated by the fact that Indiana, like anywhere else, is not a homogeneous state (especially, especially, when it comes to political primaries). It has distinct regions that differ significantly from each other and will yield dramatically different poll results. If your samples do not naturally account for these regional variations (or are not weighted to account for them), then accurate statewide primary polling is difficult and perhaps impossible.

Most polling thus far does not include crosstabs that indicate regional breakdowns. Such breakdowns are important because Democrat primary voters are effectively concentrated in four areas. These are southern Indiana, with about 40% of the votes, Indianapolis with 15%, The Region with 15%, and the North (Fort Wayne, South Bend, etc) with about 15%. The rest of the state (a vast tract of territory encompassing the Indianapolis suburbs and virtually all of central Indiana; the 3rd, 4th, 5th, and portions of the 6th and 8th districts) has only 15%.

It doesn't take a genius to look at that and determine which area is most important, at least for Democrats (the areas of importance to Republicans are roughly the inverse of this). The most recent SUSA poll counted only 26% of its respondents as being in southern Indiana, when somewhere between 35% to 40% of all Democrat primary voters will likely come from there; Hillary won southern Indiana by better than two to one.

Southern Indiana is a prime example of the inverse of the rest of Indiana, hence its dominant importance in Democratic primaries in the Hoosier state. In much of Indiana, the Republican primary is effectively the election. Accordingly, it is where voters--Republicans or not--tend to vote so as to maximize their say over their local government. Democrats in these areas tend to field incomplete slates of candidates or have unchallenged candidates in the primaries.

In Southern Indiana, the opposite is true. In many counties (though not all; Lawrence, Orange, and many others come to mind), the Democrat primary has a dominating position. Accordingly, it is where voters--Democrats or not--tend to vote in the spring (Republicans in southern Indiana have, you might say, been practicing Operation Chaos for generations). Republicans tend not to have contested primaries, and Democrats do (Harrison County is an interesting exception this year). People tend to vote in the Democrat primary as a result.

Contested primaries (regardless of location or party) tend to disproportionately draw voters, particularly when the other side's primaries are not contested.

This creates an interesting dynamic. Turnout will probably surge in the south, and not merely in places line Monroe County (Bloomington). Turnout will surge everywhere; 30+% turnout would not surprise me, nor would even near 40% turnout (by comparison, turnout in the 2004 primary was 23%).

Ohio saw statewide primary turnout surge 50% from 2004 to 2008. I suspect that we will see similar results in Indiana, with much of that turnout surge manifesting itself in the Democrat primary. For the first time in a very long time, Indiana will probably have more Democrat primary ballots pulled than Republican ones.

But this surge will be uneven. In Republican one-party counties (where the Republican primary matters, especially in central Indiana), voters will be reluctant to pull a Democrat ballot and lose a significant portion of their say over local government, regardless of what Rush Limbaugh is urging on the radio. In Democrat one-party counties (especially in southern Indiana), voters will more inclined than ever to pull a Democrat ballot.

The trend will be from Republican ballots to Democrat ones; there will be few crossovers in the other direction. Such a trend will spell the end of John McGoff's challenge to Dan Burton in the 5th district; the good doctor's candidacy always required Democrat crossover votes to have a real chance.

Most of the Democrat one-party counties are in southern Indiana. While voter turnout and Democrat primary voting will surge across Indiana, it will surge here far more than in much of central Indiana. And an Obamassiah surge to counter the Clinton surge won't come from places like Bloomington (or at least won't come enough to matter. Most of the college students will already be gone the prior weekend (finals are this week; Commencement is Saturday). This will probably negatively impact Obama's hopes of a youth surge. Finals and the move home will also erode his student volunteer base going into the final days and hours of the campaign.

Hillary Clinton will carry southern Indiana (the 8th and the 9th districts) by better than two to one, where as much as 40% of all Democrat primary votes will (if history is any indication) be cast. And she will carry the area despite Obama running close in places like Clark, Vanderburgh, Bartholomew, and perhaps Floyd. She will lose Monroe.

Southern Indiana is Hillary Clinton's bulwark in Indiana. She will probably carry the 3rd, 4th, 5th, and 6th districts by similar margins, though the numbers of Democrat primary voters in these four districts is considerably less than in the 8th and the 9th (perhaps two thirds as many, if that). The "bitter country" of the Hoosier state will go overwhelmingly for Hillary Clinton, a woman that few of those same areas would give much of a margin in November.

Obama, meanwhile, can probably count on high turnout running him up a margin of two to one in Indianapolis (the 7th District). He will also poll well among liberal yuppies in Hamilton County and in the surrounding donut counties. Even if he wins in the donut, he will not build margins there of the sort necessary to overcome the margins that Hillary will have run up elsewhere in the state. Obama will carry the Region and the 1st district (perhaps handily), though SUSA polling presently shows him only even there. The 2nd district will probably be close.

The pattern here will be a familiar one. Obama will win a few affluent and urban (and minority-heavy) counties, and perhaps those with college communities, but will lose virtually everywhere else. It will be the same pattern seen in Ohio and, more recently, Pennsylvania. The environment and the political terrain, despite being next door to Illinois, are too foreign to the Obamassiah for him to prevail; this is not Wisconsin, a state with a vibrant progressive tradition that Obama won in all demographics at the height and cresting peak of his campaign.

Hoosier Democrats are going to manifest the same buyers' remorse that blue collar and Rust Belt Democrats have shown in Ohio and Pennsylvania. Hillary will likely carry the state by about ten points (though she won't pick up very many delegates out of it).